Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update to release 24.05.7 #71

Merged
merged 30 commits into from
Mar 14, 2025
Merged

Update to release 24.05.7 #71

merged 30 commits into from
Mar 14, 2025

Conversation

itkovian
Copy link
Member

No description provided.

agilmor and others added 30 commits February 27, 2025 20:04
Cherry-pick !653 into slurm-24.05

See merge request SchedMD/dev/slurm!655
Ticket: 21836
Cherry-picked: 9babe7d
Instead of reusing the iterator variable i as the return code variable,
create a new variable rc to use as the return code. This makes reading the
code easier.

Ticket: 21836
Cherry-picked: 34e2f20
When slurmctld restarts, resv_ptr->part_ptr is set based on
resv_ptr->partition in _validate_one_reservation(). If the
resv_ptr->partition is not set, then resv_ptr->part_ptr is not set.
This can lead to a crash in certain situations where the reservation is
expected to have a non-NULL partition and part_ptr.

Ticket: 21836
Cherry-picked: 482f980
Changelog: Fix slurmctld crash when after updating a reservation with an empty
 nodelist. The crash could occur after restarting slurmctld, or if
 downing/draining a node in the reservation with the REPLACE or REPLACE_DOWN
 flag.
Ticket: 21836
Cherry-picked: 494be22
Cherry-pick !427 into slurm-24.05

See merge request SchedMD/dev/slurm!678
The list.reverse() function is an in-place operation that returns None.

Issue: 22075
Cherry-picked: b263c13
Cherry-pick !672 into slurm-24.05

See merge request SchedMD/dev/slurm!713
A nodeset with powered up and powered down nodes will be split during
node_scheduler.c:_build_node_list(). The new powered down node set needs
the weight copied from the parent node set.

Changelog: Fix jobs being scheduled on higher weighted powered down
 nodes.
Ticket: 22151
Cherry-picked: e2797cb
Cherry-pick !694 into slurm-24.05

See merge request SchedMD/dev/slurm!718
Fix a leak of gres_job_state_t.res_gpu_cores[] when
_pick_restricted_cores() is called more than once while a job is
pending.

Ticket: 22188
Changelog: Fix memory leak when RestrictedCoresPerGPU is enabled.
Cherry-picked: 9de475a
Cherry-pick !726 into slurm-24.05

See merge request SchedMD/dev/slurm!735
Continuation of commit f5027b3.

Ticket: 21266
Cherry-picked: 9ccbe69
Preparation for the next commit.

Ticket: 21266
Cherry-picked: a0a14da
job_mgr_signal_jobs() locks the assoc_mgr USER read lock. Down the call
chain, _apply_signal_jobs_filter() is called, which can call
validate_operator() if the auth_id is not the same as the job's user_id.
validate_operator() calls assoc_mgr_get_admin_level() which locks the
assoc_mgr USER read lock again.

To fix this, add a "locked" variant of validate_operator().

Ticket: 21266
Cherry-picked: f1d58f7
Changelog: Prevent slurmctld deadlock in the assoc mgr.
Ticket: 21266
Cherry-picked: c06eb30
Cherry-pick !731 into slurm-24.05

See merge request SchedMD/dev/slurm!732
Cherry-pick !737 into slurm-24.05

See merge request SchedMD/dev/slurm!740
Continuation of a0a14da

Ticket: 21266
Ticket: 22340
Cherry-picked: db16925
Cherry-pick !743 into slurm-24.05

See merge request SchedMD/dev/slurm!751
Update slurm.spec and debian/changelog as well.
Cherry-pick !747 into slurm-24.05

See merge request SchedMD/dev/slurm!749
@stdweird stdweird merged commit 4c098b7 into hpcugent:24.05.ug Mar 14, 2025
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

8 participants